ComfyUI Node: CLIP Positive-Negative XL (WLSH)

CLIP Positive-Negative XL (WLSH):

The CLIP Positive-Negative XL (WLSH) node is designed to enhance your AI art generation by leveraging the power of CLIP (Contrastive Language-Image Pre-Training) to encode both positive and negative textual prompts. This node allows you to input descriptive text that you want to emphasize (positive) and text that you want to de-emphasize or avoid (negative). By encoding these texts, the node helps in conditioning the AI model to generate images that align more closely with your creative vision. This dual conditioning approach ensures that the generated images are not only guided by what you want to see but also by what you want to avoid, providing a more refined and controlled output.

CLIP Positive-Negative XL (WLSH) Input Parameters:

clip

This parameter expects a CLIP model instance. The CLIP model is responsible for encoding the provided textual prompts into a format that can be used for conditioning the AI model. The quality and characteristics of the generated images heavily depend on the capabilities of the CLIP model used.

positive_g

This is a string parameter where you input the global positive text prompt. This text describes the elements or features you want to emphasize in the generated image. The input can be multiline, allowing for detailed descriptions. There is no strict limit on the length, but more concise prompts may yield better results.

positive_l

This is a string parameter for the local positive text prompt. Similar to positive_g, this text describes additional elements or features you want to emphasize but with a more localized focus. This can be useful for adding specific details to certain parts of the image. The input can be multiline.

negative_g

This is a string parameter where you input the global negative text prompt. This text describes the elements or features you want to avoid in the generated image. The input can be multiline, allowing for detailed descriptions. There is no strict limit on the length, but more concise prompts may yield better results.

negative_l

This is a string parameter for the local negative text prompt. Similar to negative_g, this text describes additional elements or features you want to avoid but with a more localized focus. This can be useful for removing specific details from certain parts of the image. The input can be multiline.

width

This integer parameter specifies the width of the generated image. The default value is 1024, but it can be adjusted to fit your specific needs. The minimum value is 16, and the maximum value is determined by the capabilities of your hardware and the CLIP model.

height

This integer parameter specifies the height of the generated image. The default value is 1024, but it can be adjusted to fit your specific needs. The minimum value is 16, and the maximum value is determined by the capabilities of your hardware and the CLIP model.

crop_w

This integer parameter specifies the width of the crop area. The default value is 0, meaning no cropping will be applied. Adjusting this value allows you to focus on a specific part of the generated image.

crop_h

This integer parameter specifies the height of the crop area. The default value is 0, meaning no cropping will be applied. Adjusting this value allows you to focus on a specific part of the generated image.

target_width

This integer parameter specifies the target width for the final image after any cropping or resizing. The default value is twice the width, providing a higher resolution output.

target_height

This integer parameter specifies the target height for the final image after any cropping or resizing. The default value is twice the height, providing a higher resolution output.

CLIP Positive-Negative XL (WLSH) Output Parameters:

positive

This output parameter provides the encoded positive conditioning data. It is a tuple containing the encoded positive text and additional metadata. This data is used to guide the AI model towards generating images that align with the positive prompts.

negative

This output parameter provides the encoded negative conditioning data. It is a tuple containing the encoded negative text and additional metadata. This data is used to guide the AI model away from generating images that contain elements described in the negative prompts.

positive_text

This output parameter returns the combined positive text prompts (global and local) used for encoding. It helps in verifying the exact text that was used for conditioning.

negative_text

This output parameter returns the combined negative text prompts (global and local) used for encoding. It helps in verifying the exact text that was used for conditioning.

CLIP Positive-Negative XL (WLSH) Usage Tips:

• Use concise and clear text prompts for both positive and negative inputs to achieve better results.
• Experiment with different combinations of global and local prompts to fine-tune the generated images.
• Adjust the width and height parameters to match the desired resolution of your final image.
• Utilize the crop parameters to focus on specific areas of the generated image if needed.

CLIP Positive-Negative XL (WLSH) Common Errors and Solutions:

"Invalid CLIP model instance"

• Explanation: The clip parameter does not contain a valid CLIP model instance.
• Solution: Ensure that you are passing a properly initialized CLIP model to the clip parameter.

"Text prompt too long"

• Explanation: The provided text prompt exceeds the maximum length that the CLIP model can handle.
• Solution: Shorten the text prompt to fit within the model's maximum token limit.

"Invalid width or height value"

• Explanation: The width or height parameter is set to a value outside the acceptable range.
• Solution: Ensure that the width and height values are within the specified minimum and maximum limits.

"Cropping dimensions exceed image dimensions"

• Explanation: The crop width or height exceeds the dimensions of the generated image.
• Solution: Adjust the crop width and height to be within the dimensions of the generated image.